AutoCore与研华携手高通,于CES推出下一代机器人平台

该软硬一体化平台基于高通 Dragonwing™ IQ-9075™ 平台、研华机器人控制系统 AFE-A503 以及 AutoCore 的机器人软件平台 AutoCore.OS™,旨在加速智能机器人从原型到商业化部署的进程。

2026-01-05
AutoCore

[拉斯维加斯· CES] — [2026年1月5日] 领先的智能移动出行与机器人平台解决方案提供商 ——AutoCore 与研华( Advantech )今日在国际消费类电子产品展览会( CES )上宣布,将与高通技术公司( Qualcomm Technologies, Inc. )合作构建并推出下一代智能机器人平台。该平台深度整合了研华基于 高通Dragonwing™ IQ-9075 处理器 的工业级机器人硬件平台,以及 AutoCore 具备量产能力的机器人软件平台 AutoCore.OS™ 。其目标是帮助全球机器人开发者和企业消除工程瓶颈,加速实现从创新原型到大规模商业化部署的跨越。

在将创新原型转化为可规模部署的产品过程中,许多机器人项目常被通信不稳定、软硬件集成难度大、OTA(在线升级)与安全能力缺失等“最后一公里”的工程问题所阻滞。AutoCore与研华和高通的合作,旨在通过提供一个预集成、预验证的软硬件一体化解决方案,根除这些障碍。

研华科技总裁 Miller 表示:“在研华,我们看到机器人技术正迈入一个新阶段 —— 成功的定义不再仅仅取决于原型的性能表现,而在于智能机器人如何可靠、高效地规模化部署到真实世界中。通过与高通技术公司和AutoCore 的合作,并依托我们的 AFE-A503 机器人控制平台,我们正在交付一个工业级、为量产而生的基座。这使得机器人开发者能够专注于创新,而我们则负责保障从原型到大规模部署过程中的鲁棒性、全生命周期管理以及可扩展性。”

“我们与研华和高通的合作,旨在为机器人行业提供一个坚实、可信赖的基座。我们的目标是把客户从底层的无差别的工程复杂性中解放出来,让他们能百分之百地专注于产品差异化与核心算法的创新。” AutoCore 首席执行官张旸说道。

“高通技术公司致力于通过我们开创性的 AI 和连接技术,赋能机器人领域的创新,”高通技术公司产品管理副总裁 Anshuman Saxena 表示,“我们与研华(Advantech)和 AutoCore 的合作,将我们先进的高通 IQ-9075 平台与他们可商用的硬件及量产级的软件栈相结合,为开发者社区提供了一个强大的工具,能够显著加速下一代机器人的开发与部署。“

基于 AFE-A503 的具身机器人整机参考设计

为直观展示 AFE-A503 平台的强大能力,AutoCore 开发了一个具身机器人整机参考设计。该机器人基于融合了AutoCore软件平台的 AFE-A503 软硬一体化平台,集成了一个轮式底盘和两支多自由度机械臂,它能够:

  • 自主导航与避障:在复杂的模拟环境中,自主规划路径并灵活避开动态障碍物
  • 目标识别与定位:通过AI视觉算法,精准识别并定位指定的目标物体(如桌上的水瓶)。
  • 协同抓取作业:移动底盘将机器人运送至最佳抓取位置,随后机械臂精准地将目标物体抓起并放置到指定区域。

这个演示生动地诠释了AFE-A503的愿景:通过提供一个开箱即用的强大平台,使开发者能够在数周而非数月内快速构建出功能完整的智能机器人原型,从而将宝贵的研发资源聚焦于上层应用与算法创新,极大加速产品上市进程。

研华科技(Advantech)核心能力 – AFE-A503 机器人解决方案

随着智能机器人从实验验证迈向关键任务部署,行业亟需能够将AI性能、系统行为与运行可靠性完美融合的集成化解决方案。研华AFE-A503正是这样一款以解决方案为导向的机器人平台,它集成了计算能力、系统架构与工业级设计,旨在支持可扩展的、真实场景下的机器人应用。

  • 基于高通Dragonwing™ IQ-9075的高性能机器人计算平台
    AFE-A503由高通 Dragonwing™ IQ-9075 平台赋能,通过结合高性能CPU、GPU 和 AI加速器的异构计算架构,提供高达100 TOPS的 AI 性能。这使得感知、规划和控制工作负载能够在单个机器人控制器上并发运行,支持边缘端复杂的AI推理与实时决策。
  • 以机器人为中心的系统架构
    AFE-A503 专为机器人设计,支持混合实时与AI工作负载、多传感器融合,以及针对视觉摄像头、激光雷达(LiDAR)和机器人外设的高带宽连接。其系统架构针对低延迟数据通路和确定性行为进行了优化,提供了先进智能机器人所需的响应速度与灵活性。
  • 面向规模化部署的工业级设计
    AFE-A503 采用工业级供电设计、散热架构及系统验证,确保在持续的 AI 工作负载和连续运行场景下保持稳定运行。凭借长生命周期的供货承诺与全球供应链支持,它助力实现从原型开发到大规模商业化机器人部署的平滑过渡。
  • 平台级赋能的研华机器人套件(Advantech Robotic Suite)
    构建在AFE-A503硬件基础之上,研华机器人套件提供了一套机器人中间件、驱动程序、SDK 和集成工具,简化了系统启动、传感器集成及AI流水线部署的过程。该套件与 AutoCore.OS及ROS2生态系统协同工作,加速开发流程,帮助机器人开发者高效地从原型验证迈向可商用的部署阶段。
AutoCore核心能力 — 为规模化量产而生的全栈平台

AutoCore将复杂的底层技术栈封装为简单易用的平台服务,让开发者无需深陷泥潭,即可构建稳定、高效、安全的机器人应用。

  • 量产级机器人软件平台:AutoCore.OS
    AutoCore.OS基于创新的“双内核”异构计算架构,在单一芯片上实现了确定性实时任务(如毫秒级运动控制)与高吞吐非实时任务(如 AI 视觉、多传感器融合)的完美隔离与高效协同。在内核层,通过引入实时内核,并内置AutoCore EtherCAT主站协议,为多轴伺服系统提供微秒级的控制;在非实时域基于Linux内核,实现跨域高效IPC机制,并提供完整的机器人基础软件栈,包括商业级高性能通信协议栈,并注入机器人从原型开发到商业化量产落地所必需的核心能力,包括系统级健康管理机制,应用程序运行时管理框架,系统诊断,数据加解密服务,系统休眠唤醒机制,参数配置和标定机制,轻量化资源隔离和访问控制的沙盒(Sandbox)以及软件在线升级(OTA)等,在完全兼容ROS2系统的同时,从根本上解决了基于开源ROS2方案在复杂网络环境,复杂应用场景下的通信抖动与不确定性,复杂协作任务的调度性能不佳,系统稳定性差等核心问题,完美填补了使用开源ROS2从原型到量产之间的鸿沟,为机器人商业化落地提供可预测、可管理、高可靠的基础底座。
  • 加速应用开发与部署的软件栈
    该平台还提供了一系列开箱即用且深度优化的机器人算法和框架模块。它内置了经过量产验证的 Autoware与Nav2 导航算法包,并原生集成了MoveIt! 框架。开发者无需从零开始,即可快速实现高精度的自主导航与复杂的多轴机械臂协同控制,从而将精力聚焦于业务逻辑与场景创新,实现“开箱即用,快速上市”。
关于AutoCore

AutoCore是全球领先的智能移动平台解决方案提供商,自2018年成立以来,始终致力于为智能汽车与机器人产业提供可量产、可扩展、高安全的功能安全软件平台。作为国家级专精特新“小巨人”企业,AutoCore专注于解决从原型到商业化部署过程中的核心挑战,帮助合作伙伴加速产品上市进程。

关于高通

高通公司是全球领先的无线科技创新者,也是5G研发、商用与实现规模化的推动力量。把手机连接到互联网,我们的发明开启了移动互联时代。今天,我们的基础科技赋能了整个移动生态系统,每一台3G、4G和5G智能手机中都有我们的发明。我们正在将移动技术的优势带到汽车、物联网、计算等全新行业,开创人与万物智能互联的新世界。

关于研华

研华科技是物联网智能系统与嵌入式平台的全球领导厂商,以“智能地球的推手”为企业愿景。为迎接边缘计算与人工智能的趋势,研华全面部署产业驱动战略,聚焦于边缘计算与边缘AI领域。同时,研华通过整合其边缘计算硬件平台、WISE-IoT软件平台,以及结合领域专业知识的特定行业边缘AI解决方案,不断增强其全球布局与核心竞争力。

AutoCore and Advantech Collaborates with Qualcomm to Launch Next-Generation Robotics Platform at CES

Based on the Qualcomm Dragonwing™ IQ-9075TM platform and Advantech Robot Control System AFE-A503 and AutoCore’s robotics software platform AutoCore.OSTM, the integrated hardware-software platform aims to accelerate intelligent robots from prototype to commercial deployment.

[Las Vegas · CES] — [January 5th, 2026] Advantech and AutoCore, the leading providers of smart mobility and robotics platform solutions, today announced at the Consumer Electronics Show (CES) a collaboration with Qualcomm Technologies, Inc. to build and launch a next-generation intelligent robotics platform. This platform deeply integrates Advantech industrial grade robotics hardware platform based on the Qualcomm Dragonwing™ IQ-9075 processor—with AutoCore’s production-ready robotics software platform, AutoCore.OSTM. It aims to help global robot developers and enterprises eliminate engineering bottlenecks and accelerate the leap from innovative prototypes to large-scale commercial deployment.

In the process of transforming innovative prototypes into scalable products, many robotics projects are often stalled by "last-mile" engineering problems such as unstable communication, difficulty in hardware-software integration, and a lack of OTA (Over-The-Air) and security capabilities. The collaboration between AutoCore and Qualcomm aims to resolve these obstacles by providing a pre-integrated, pre-verified hardware-software integrated solution.

“At Advantech, we see robotics entering a new phase—where success is no longer defined by prototype performance, but by how reliably and efficiently intelligent robots can scale into real-world deployment,” said Miller, President at Advantech. “By collaborating with Qualcomm Technologies and AutoCore, and through our AFE-A503 robotics control platform, we are delivering an industrial-grade, production-ready foundation that allows robot developers to focus on innovation, while we take care of robustness, lifecycle, and scalability from prototype to mass deployment.”

“Our collaboration with Advantech and Qualcomm Technologies aims to provide a solid, trustworthy foundation for the robotics industry. Our goal is to liberate customers from undifferentiated underlying engineering complexities, allowing them to focus 100% on product differentiation and core algorithmic innovation," said Yang Zhang, CEO of AutoCore.

"Qualcomm Technologies is committed to empowering innovation in robotics through our pioneering AI and connectivity technologies,” stated Anshuman Saxena, Vice President, Product Management, Qualcomm Technologies, Inc. “Our collaboration with Advantech and AutoCore combines our advanced Qualcomm IQ-9075 platform with their commercial ready hardware and production-grade software stack, providing the developer community with a powerful tool that significantly accelerates the development and deployment of next-generation robots."

Reference Design for Complete Embodied Robot Based on AFE-A503

To visually demonstrate the powerful capabilities of the AFE-A503 platform, AutoCore has developed a reference design for a complete Embodied robot. Based on the AFE-A503 integrated hardware-software platform, the robot integrates a wheeled chassis and two multi-degree-of-freedom robotic arms, capable of:

  • Autonomous Navigation and Obstacle Avoidance: Autonomously planning paths and flexibly avoiding dynamic obstacles in complex simulated environments.
  • Target Recognition and Localization: Accurately recognizing and locating specific target objects (such as a water bottle on a table) using AI vision algorithms.
  • Collaborative Grasping Operations: Moving the chassis to the optimal grasping position, followed by the robotic arms precisely grasping the target object and placing it in a designated area.

This demonstration vividly interprets the vision of AFE-A503: by providing a powerful, out-of-the-box platform, developers can quickly build fully functional intelligent robot prototypes in weeks rather than months, thereby focusing valuable R&D resources on upper-layer applications and algorithmic innovation, greatly accelerating time-to-market.

Advantech Capabilities – AFE-A503 Robotics Solution

As intelligent robots transition from experimentation to mission-critical deployment, the industry demands integrated solutions that align AI performance, system behavior, and operational reliability. Advantech AFE-A503 represents a solution-oriented robotics platform that brings together computing, system architecture, and industrial-grade design to support scalable, real-world robotic applications.

  • High-Performance Robotics Compute on Qualcomm Dragonwing™ IQ-9075
    Powered by the Qualcomm Dragonwing™ IQ-9075 platform, AFE-A503 delivers up to 100 TOPS of AI performance through a heterogeneous computing architecture combining high-performance CPU, GPU, and AI acceleration. This enables perception, planning, and control workloads to run concurrently on a single robot controller, supporting complex AI inference and real-time decision-making at the edge.
  • Robotics-Centric System Architecture
    Designed specifically for robotics, AFE-A503 supports mixed real-time and AI workloads, multi-sensor fusion, and high-bandwidth peripheral connectivity for vision cameras, LiDARs, and robotic peripherals. Its system architecture is optimized for low-latency data paths and deterministic behavior, providing the responsiveness and flexibility required for advanced intelligent robots.
  • Industrial-Grade Design for Scalable Deployment
    Built with industrial-grade power delivery, thermal architecture, and system validation, AFE-A503 ensures stable operation under sustained AI workloads and continuous operation scenarios. With long lifecycle availability and global supply support, it enables a smooth transition from prototype development to large-scale, commercial robot deployment.
  • Advantech Robotic Suite for Platform-Level Enablement
    Building on the AFE-A503 hardware foundation, Advantech Robotic Suite provides a set of robotics middleware, drivers, SDKs, and integration tools that simplify system bring-up, sensor integration, and AI pipeline deployment. Working alongside AutoCore.OS and the ROS2 ecosystem, Robotic Suite accelerates development workflows and helps robot developers move efficiently from prototype validation to commercial-ready deployment.
AutoCore Core Capabilities — A Full-Stack Platform for Scalable Mass Production

The AutoCore software platform encapsulates complex underlying technology stacks into simple, easy-to-use platform services, allowing developers to build stable, efficient, and secure robot applications without getting bogged down in technical details.

  • Production-Grade Robotics Software Platform: AutoCore.OS
    AutoCore.OS is based on an innovative "dual-kernel" heterogeneous computing architecture, achieving perfect isolation and efficient collaboration between deterministic real-time tasks (such as millisecond-level motion control) and high-throughput non-real-time tasks (such as AI vision and multi-sensor fusion) on a single chip. At the kernel layer, by introducing a real-time kernel and incorporating the AutoCore EtherCAT Master protocol, it provides microsecond-level control for multi-axis servo systems; in the non-real-time domain based on the Linux kernel, it implements an efficient cross-domain IPC mechanism and provides a complete robotics infrastructure software stack. This includes a commercial-grade high-performance communication protocol stack and injects core capabilities essential for robots from prototype development to commercial mass production, including system-level health management mechanisms, application runtime management frameworks, system diagnostics, data encryption/decryption services, system sleep/wake mechanisms, parameter configuration and calibration mechanisms, lightweight resource isolation and access control sandboxes, and software Over-The-Air (OTA) upgrades. While remaining fully compatible with the ROS2 system, it fundamentally resolves core issues found in open-source ROS 2 solutions under complex network environments and application scenarios—such as communication jitter and uncertainty, poor scheduling performance for complex collaborative tasks, and poor system stability—perfectly bridging the gap between prototype and mass production using open-source ROS2, and providing a predictable, manageable, and highly reliable foundation for commercial robot deployment.
  • Software Stack Accelerating Application Development and Deployment
    The platform also provides a series of out-of-the-box and deeply optimized robotics algorithms and framework modules. It features built-in, production-verified Autoware and Nav2 navigation algorithm packages and natively integrates the MoveIt! framework. Developers can quickly implement high-precision autonomous navigation and complex multi-axis robotic arm collaborative control without starting from scratch, allowing them to focus on business logic and scenario innovation, achieving "out-of-the-box readiness and rapid time-to-market."
About AutoCore

AutoCore is a global leading provider of smart mobility platform solutions. Since its establishment in 2018, it has provided production-ready, scalable, and highly secure functional safety software platforms for the intelligent automotive and robotics industries. As a national-level specialized and sophisticated "Little Giant" enterprise, AutoCore focuses on solving core challenges from prototyping to commercial deployment, helping partners accelerate product time-to-market. The company's core product, AutoCore.OS, is the industry's first full-stack software platform to receive ISO 26262 ASIL-D functional safety product certification. It adapts to heterogeneous distributed computing architectures and various chip platforms, supports multiple operating systems, and provides developers with comprehensive toolchain support, offering a solid and reliable foundation for Software-Defined Vehicles (SDV) and intelligent robots.

About Qualcomm

Qualcomm is the world’s leading wireless technology innovator and the driving force behind the development, launch, and expansion of 5G. When we connected the phone to the internet, the mobile revolution was born. Today, our foundational technologies enable the mobile ecosystem and are found in every 3G, 4G, and 5G smartphone. We bring the benefits of mobile to new industries, including automotive, the internet of things, and computing, and are leading the way to a world where everything and everyone can communicate and interact seamlessly.

About Advantech

Advantech is a global leader in IoT intelligent systems and embedded platforms, with the corporate vision of "Enabling an Intelligent Planet." To embrace the trends of edge computing and artificial intelligence, Advantech fully deploys its sector-driven strategy and focus on Edge Computing and Edge AI. In the meantime, Advantech is enhancing its global presence and core competitiveness by integrating its Edge Computing hardware platform, the WISE-IoT software platform, and sector-specific Edge AI solutions with domain expertise.